Heritage Language Learning and the Portuguese Subjunctive* Gláucia V. Silva

University of Massachusetts Dartmouth

Abstract: In a study of the perception and production of the subjunctive, heritage and non-heritage learners of Portuguese completed four different types of written tasks, besides a questionnaire to determine their linguistic background and some of their beliefs about their own language abilities. Participants were enrolled in college language courses, in three different levels: high beginner, intermediate and low advanced. The results obtained show that heritage learners produce more accurate verb forms than do non-heritage learners, although the latter increase their accuracy notably in the more advanced levels. Since heritage learners produce a high number of accurate forms even in lower levels, it is suggested that differentiated instruction may be in the best interest of both heritage and non-heritage learners of Portuguese. Keywords: Heritage learners, Portuguese, grammar production, subjunctive, indicative

Since the early 19th century, the Northeast region of the United States has received many

immigrants from Portugal, both from the continent and from the islands (Azores and Madeira).

By the mid-1800s, there was a “significant and stable” Portuguese community in the U.S.

(Stephens). Although immigration rates from Portugal to the U.S. have decreased considerably in

the past two to three decades, the American Northeast has seen more recent immigration from

other Portuguese-speaking countries—notably, Brazil and Cape Verde1 (Ferreira, “That’s not”;

Jouët-Pastré). With a sizeable Portuguese-speaking population, Portuguese language classes in

the American Northeast tend to be attended by learners who have been exposed to the language

from an early age—i.e., by heritage learners.

Heritage language (HL) learning is a field that has gained prominence in the U.S. in the

last three decades. Valdés et al’s seminal volume on Spanish heritage language teaching called

attention to the fact that HL learners have needs that are different from those of non-heritage

language (NHL) learners. These needs have been the focus of much of the research into HL

learning (Acevedo; Fairclough; Kondo-Brown “Differences”, Heritage Language; Kondo-Brown

2

and Brown; Tse; Valdés et al; among others). This study contributes to that discussion by

analyzing the perception and the production of a particular grammar feature, the subjunctive

mood, by heritage and non-heritage language learners of Portuguese in Southeastern

Massachusetts. Specifically, this paper investigates whether Portuguese HL learners differ from

NHL learners in their perception and production of the subjunctive in written tasks. It also

examines whether length of formal instruction affects the production of target-like forms in

written tasks, and whether there is a correlation between self-assessment and actual production of

the subjunctive in the three different levels. After an analysis of the data obtained, a discussion is

offered on what the findings may signify for HL teaching and learning.

Background and Motivations for This Study

As mentioned above, the American Northeast has a sizeable population of immigrants

from Portuguese-speaking countries, and college Portuguese language classes in Southeastern

Massachusetts are attended by significant numbers of heritage learners (mostly of European

descent at this point). In spite of the large number of Portuguese heritage learners in

Massachusetts and elsewhere in the U.S. (especially California, New York, New Jersey and

Florida), few studies have investigated this population and their needs in terms of language

learning. Notable exceptions are studies by Ferreira (“That’s Not”, “Portuguese”), which address

sociolinguistic characteristics displayed by her students of Portuguese.

Valdés (“The Teaching”) and others have stated that heritage learners may not have been

exposed to the standard variety of their HL, since the use of the HL is restricted to certain

contexts. Therefore, heritage learners may face some difficulties when presented with a language

variety that does not include forms that they can readily recognize. English is the language of

3

education for these speakers, while the HL is the variety spoken at home—a variety that may not

be the standard for that language. According to several authors, the goal of HL teaching needs to

be bidialectalism: teachers must recognize and validate the learners’ variety, while teaching the

standard dialect of the HL (Bernal-Enríquez and Hernández Chávez; Fairclough “La

adquisición,” Spanish; Gutiérrez Marrone; Valdés “Implications”; among others). Teachers

would then show that the variety that learners speak/understand is valid in certain situations,

whereas other situations (such as professional environments) might require the use of the variety

that is considered standard. That way, learners would become bidialectal, having learned one

more dialect (the standard) of their HL.

The restricted social use of heritage languages leads to a system which contains

innovative (i.e., non-normative) elements (Lynch, “Similarities” 253). Moreover, contact with

English affects the variety acquired by HL learners: although the lexicon may constitute the most

observable evidence of language contact2, the verb system may also be affected by contact with

another language (as argued, for Spanish, by Silva-Corvalán 167).

In Romance languages, the subjunctive is a mood that is morphologically marked. In

English, on the other hand, the subjunctive appears only residually, as in (1) below:

(1) a. If I were you, I’d do the work. b. It is important that she be there.

The forms in italics in (1a) (were) and (1b) (be) illustrate uses of the subjunctive in English. In

informal spoken English, however, it may be just as common (if not more so) to find sentences

such as “If I was you” and “It is important for her to be there” (or some such variation), in which

the subjunctive is no longer utilized. In Portuguese, on the other hand, the subjunctive is

generally quite productive3.

4

But what is verb mood? And when is the subjunctive mood used in Portuguese? Perini

explains that mood is “a mixed category having to do with both meaning and form” (167).

According to Perini, differences in mood may account for differences in meaning, but a certain

verb mood may also be required by the presence of a governing lexical item. In terms of

meaning, it has been noted that mood is related to the speaker’s perspective: for example, use of

the subjunctive may signify that an event is unexperienced, uncertain, unproven (Bull). More

recently, it has been argued that the use of the subjunctive is associated with the notion of

relevance (Lunn; Ahern and Leonetti): the indicative mood marks relevant information, what the

speaker asserts; the subjunctive, on the other hand, is used for non-assertion, for information

deemed of relatively low value by the speaker. Although Bull’s, Lunn’s, and Ahern and

Leonetti’s claims were originally made for Spanish, they may be applied to Portuguese as well.

The sentences in (2) exemplify uses of the three simple tenses of the subjunctive in

Portuguese (present in (2a), past in (2b) and future in (2c)):

(2) a. Eu prefiro que ela vá à reunião. I prefer that she go to the meeting. b. Se ela fosse à reunião, nós chegaríamos a um acordo. If she went to the meeting, we would reach an agreement. c. Quando ela for às reuniões, resolveremos tudo. When she goes to the meetings, we will solve everything.

The examples in (2) illustrate unexperienced events (as in (2a)) and uncertain events (as in (2b)

and (2c)). Furthermore, a verb such as preferir, “to prefer” governs the mood of the verb in the

subordinate clause: while the main verb preferir (as in (2a)) appears in the indicative, the

subordinate verb (in (2a), the verb ir, “to go”) must occur in the subjunctive. In (2b) and (2c) the

subjunctive conveys that the event may or may not happen (and in (2b) there is even less

5

certainty that the event might happen). If the indicative were used in those examples, the hearer

would know that sometimes she does go to the meetings and they do reach an agreement4.

In terms of acquisition of L1, the subjunctive may appear late, as argued by Blake for

Spanish. Gragera notes that features that appear later in the acquisition process are the first to

weaken if the speaker is immersed in a foreign language environment, as is the case with both

HL and NHL learners. Moreover, Lynch (“Relationship”) notes that NHL and HL learners have

more problems with forms that are less frequent in the language or that appear later in the

acquisition process. Hislope reports on a study that focused on production of present subjunctive

by heritage speakers of Spanish. She found that production of this verb tense was sporadic,

which may be expected if a form is undergoing simplification in the language. Establishing a

parallel with Spanish, we may assume that the subjunctive appears late for children acquiring

Portuguese. If this is the case, we would expect that learners would have trouble with this mood

in Portuguese, as Hislope has found for Spanish.

My experience seems to corroborate what Hislope notes for HL learners of Spanish.

Several Portuguese HL learners seem to exhibit difficulties with uses of the subjunctive and the

indicative: sometimes, the indicative is used for the subjunctive, as in (3); other times, one form

of the subjunctive is used when another would appear in standard (European or Brazilian)

Portuguese, as in (4) (the corresponding standard forms are given in (3’) and (4’)):

(3) Se as pessoas podiam, moravam no campo. (3’) Se as pessoas pudessem, moravam no campo. If people could, they would live in the country. (4) O que vocês querem fazer hoje quando nós cheguemos no Boston? (4’) O que vocês querem fazer hoje quando nós chegarmos a Boston? What do you(pl) want to do today when we arrive in Boston?

6

The type of production exemplified in (3) and (4) are at the root of this study, which

looks at whether and how heritage learners differ from foreign language learners regarding the

perception and production of subjunctive and indicative forms, and whether the perception and

production of the subjunctive is affected by length of formal instruction. The next section briefly

outlines the design of the study.

Project Design

A total of 56 learners took part in this study. Participants were college students enrolled

in Portuguese language classes in Southeastern Massachusetts. The classes were taught by three

different instructors: a native speaker of Brazilian Portuguese, a native speaker of European

Portuguese, and a non-native speaker of European Portuguese. Students filled out a learner

profile/self-evaluation questionnaire designed to assess their background (heritage or foreign

language learners) and determine level of instruction. The profile questionnaire also contained

statements about grammar and about the subjunctive that were assessed in a simplified Likert-

type scale. Appendix A contains the learner profile/self-evaluation questionnaire. Learners were

also asked to complete four different tasks designed to elicit perception and production of verb

forms (indicative vs. subjunctive) in Portuguese. The four tasks were: a multiple choice exercise

with eight questions, a cloze-type exercise with 30 tokens, a short paragraph, and a

grammaticality judgement task with 10 sentences. The perception and production tasks, written

in standard European Portuguese, can be found in Appendix B. The design of the questionnaire

and of the tasks was inspired by those presented in Santos and Silva and Fairclough (Spanish).

The questionnaire and the tasks were completed anonymously during class time. For

matching purposes, participants received a number to write on top of the questionnaire and the

7

task sheets. On average, learners took approximately 25 minutes to complete the questionnaire

and tasks. Participants were enrolled in classes that ranged from high beginning to low advanced

levels. The next two sections discuss the results obtained in the background/self-perception

questionnaire and in the perception/production tasks.

Learner Profiles and Self-Assessment

In this study, most participants (71.5%) were heritage learners of European Portuguese.

As mentioned above, for the purposes of this study a heritage learner was someone who was

exposed to Portuguese when s/he was growing up. In other words, participants who included

Portuguese in the answers for questions 1 and/or 2 of the learner profile questionnaire (see

Appendix A) were considered heritage learners.

In terms of enrollment, 41% of learners were enrolled in low advanced classes. The

second largest group (34%) were enrolled in high beginning classes. The remaining participants

(25%) were taking classes at the intermediate level. All three groups had had some formal

instruction on the subjunctive, but learners in high beginning classes only received instruction on

each tense of the subjunctive as it appeared in the textbook, which only addressed each tense

once. Uses of the subjunctive were discussed and practiced at least one more time (over several

class meetings) in each of the two other groups.

Overall, respondents (both HL and NHL learners) felt comfortable or somewhat

comfortable with their grammar skills. Nevertheless, learners were divided about their own

production of verbs in Portuguese: while most NHL learners (62.5%) agreed with Statement (c)

(“I have problems with different verb tenses in Portuguese”), HL learners were split between

agreeing and disagreeing with the statement: 45% agreed that they had problems, while 35% did

8

not agree (the remaining 20% neither agreed nor disagreed). However, when looking at the

results by level, there appears to be some correlation between level and answer for HL learners:

the higher the level, the less HL learners agree with Statement (c). While 64% of learners in the

high beginning level admitted to general problems with verb tenses, only less than half the

learners in the low advanced level thought that verb tenses in general were problematic. This

correlation does not exist for NHL learners.

Although most NHL learners stated that they had problems with verbs in general, only

25% of learners in this group agreed that the subjunctive is the most difficult verb form in

Portuguese (Statement (d)). As a group, HL learners did not think that the subjunctive stood out

in difficulty either. At first glance, these results appear to indicate that learners in general do not

perceive the subjunctive as a particularly challenging grammar point. This preliminary

conclusion is warranted by further analysis of the results: learners in both groups appear to be

fairly confident in their abilities to employ the subjunctive in written language, as evidenced by

responses to Statement (g): in both groups, over 40% of learners agreed that they know when to

use the subjunctive in written language. However, it must be noted that NHL learners were more

confident about the use of this verb form in written than in spoken language: the majority of

NHL learners (over 55%) neither agreed nor disagreed that they know how to use the subjunctive

in spoken language. Finally, the two groups of learners exhibit differences regarding whether

they consider their speaking abilities in Portuguese better than their writing abilities. This is

shown in the results for Statement (e) (“I speak Portuguese better than I write it”). Not

surprisingly, HL learners tended to agree with (e), while most NHL learners disagreed with the

statement. Figures 1 and 2 show the results of the self-assessment questionnaire for HL and NHL

learners, respectively.

9

Figure 1: Self-assessment by HL learners.

0

10

20

30

40

50

a) b) c) d) e) f) g) h)

statements

%

AgreeA/DDisagree

Figure 2: Self-assessment by NHL learners.

010203040506070

a) b) c) d) e) f) g) h)

statements

%

AgreeA/DDisagree

Perception and Production Tasks

As mentioned above (see Project Design), this study included four written tasks, two that

can be considered perception tasks (a multiple choice exercise and a grammaticality judgement

task) and two production exercises (a cloze-type exercise and a short paragraph).

Multiple Choice Task

The multiple choice exercise contained 8 sentences with one verb left blank (see

Appendix B). Learners were given three choices to complete the sentence. In standard dialects of

Portuguese, all blanks corresponded to a form of the subjunctive (either present, past or future).

10

Results of this task indicate that HL learners recognize uses of the subjunctive more often than

do NHL learners. The difference between the performance of HL and NHL learners was more

pronounced for learners enrolled in lower level classes, as seen in Figure 3:

Figure 3: Accuracy for multiple choice task.

010203040506070

High Beg Interm Low Adv

%

HLFL

Cloze-type task

The general results for the cloze-type exercise also indicate that NHL learners progress

consistently over time regarding uses of the subjunctive and the indicative moods. In this task,

the difference between the performance of the two groups is much smaller than in the multiple

choice task. The biggest difference between the groups is found in the intermediate level, where

HL learners performed with 74.2% accuracy, while NHL learners gave target-like responses in

61.7% of the cases. In the other two levels, HL learners performed only slightly better than NHL

learners: the differences are 1% for the high beginner group and 2% for the low advanced group.

Figure 4 illustrates the performance of HL and NHL learners in the cloze task.

11

Figure 4: Accuracy in cloze-type exercise.

0102030405060708090

High Beg Interm Low Adv

%

HLFL

The inaccuracies in the cloze exercise were not all cases of indicative used when the

subjunctive was expected and vice-versa. We also found that learners may use the subjunctive or

the indicative in a tense that is different from the target form. This is the case, for example, in (5)

(which corresponds to blank 29 in the exercise—see Appendix B):

(5) Se o tempo esteja mau, (Pedro) ficará em casa. If the weather is (pres. subj.) bad, Pedro will stay at home.

In (5), the target form of the verb estar is the future subjunctive (estiver). The present

subjunctive was produced by learners in all three levels, and both by HL and by NHL learners.

This type of production indicates that both HL and NHL learners may realize that the subjunctive

should be used in certain cases (such as hypotheticals), even though they may not know which

form of the subjunctive to use. The same type of mix-up is also found in the indicative mood, as

exemplified in (6) (blank 20 in the exercise):

(6) …mas a irmã ainda não descobre e talvez não descubra nunca. ...but the sister still has not found out (pres. indic.) and maybe won’t ever find out.

12

The target form of the verb descobrir (“to find out”) in (6) would be the preterite (descobriu), but

some learners (both in the HL and in the NHL groups) produced the present and not the past

form of the verb.

In the cloze-type exercise, we also found inaccuracies related to spelling: in some cases,

learners appeared to know which form to use, but not how to spell them, as shown in (7):

(7) Se eles dizerem que vai fazer sol, ele vai ao norte. If they say that it will be sunny, he will go to the north.

In (7), the verb dizer (“to say”) should appear in the future subjunctive, disserem. A few learners

in all three levels (both HL and NHL learners) spelled the verb with a z instead of ss. This

example is noteworthy because this mistake transcends a mere spelling error: the form dizerem

also exists in Portuguese, but it is a form of the inflected (or personal) infinitive. The relevance

of (7) lies in the fact that learners may have been exposed to the form that they produced and

were confused by the close spelling.5

The production found in the cloze-type exercise also includes some more trivial spelling

errors, such as fassa for faça (“do” in present subjunctive, first or third person singular) by a high

beginning HL learner; chouver for chover (“rain” in future subjunctive, third person singular), by

an intermediate NHL learner; and descobreu for descobriu (“discover” in simple preterite

simple, third person singular), by a low advanced HL learner. Other spellings reveal what the

learner had heard before receiving formal instruction, as is the case of oiça (“hear” in present

subjunctive, first or third person singular). Although the textbooks used by the learners surveyed

have only the form ouça for the present subjunctive of the verb ouvir (“to hear”), the form oiça is

also found in European Portuguese, and was produced by a few HL learners. Note that oiça was

considered accurate in the coding of the data.

13

Short Paragraph

The short paragraph task intended to elicit forms of the present subjunctive. For that end,

the prompt was O que esperas do próximo Presidente?, “What do you expect from the next

President?” There was no minimum or maximum number of words specified in this task. As a

result, some paragraphs were rather short. The coding was done by counting how many verb

forms were accurate indicative, inaccurate indicative, accurate subjunctive, and inaccurate

subjunctive forms in each level (separating HL and NHL learners). Then the percentage of verb

forms represented in each category was calculated. Figure 5 shows the percentages of accurate

indicative and subjunctive use by HL and NHL learners in all three levels (where HB stands for

High Beginning, Int for Intermediate, and LA for Low Advanced):

Figure 5: Accuracy in paragraph task.

01020304050607080

HBIndic

HBSubj

IntIndic

Int Subj LAIndic

LA Subj

%

HLFL

The results for the paragraph task indicate that in all three levels, the HL group produced

more accurate forms of the subjunctive than did the NHL group. Note that the NHL group

produced proportionately more accurate forms of the indicative than did the HL group in the

three levels. At first glance, this might suggest that the NHL group is more proficient in the

indicative mood than learners in the HL group. Closer examination of the data, including the

incorrect tokens, shows that NHL learners overall used the subjunctive much less than did HL

14

learners: while 34.4% of the verbs used by the HL group were in the subjunctive mood, only

20.2% of the tokens produced by NHL learners were a form of the subjunctive. Thus, we find

among NHL learners a tendency to rely on the indicative mood, which may account for the

higher number of accurate forms. Naturally, this tendency also correlates with the higher number

of accurate forms of the subjunctive found in the HL group. Therefore, a relevant result of the

analysis of the paragraph task is the fact that HL learners produce proportionately more tokens of

the subjunctive than do NHL learners.

Analysis of the forms produced in the paragraph task reveals that HL learners may use

both the subjunctive and the indicative in structures where the subjunctive is the target form, as

illustrated in examples (8-10):

(8) Espero que os próximos presidentes ajudam o país. […] Espero que eles façam mais […]. I hope that the next presidents help (pres. indic.) the country. I hope that they do (pres. subj.) more. (High Beginner level) (9) Eu espero que o próximo presidente termina a guerra e traga as nossas tropas [sic] […]. I hope that the next president ends (pres. indic.) the war and brings (pres. subj.) our troops… (Intermediate level) (10) Eu espero que o próximo presidente seja melhor. [...] Mais importante é que ele acaba com a guerra. I hope that the next president is (pres. subj.) better. More important is that he ends (pres. indic.) the war. (Low Advanced level)

Among NHL learners we found some inconsistency regarding use of tense of the

subjunctive in the same type of structure. In (11), both verbs in italics should be in the present

subjunctive, but the verb considerar (“to consider”) appears in the past tense, while the verb

tomar (“to take”) is in the target form:

15

(11) Espero que ele considerasse os desejos do público e que não tome tão [sic] dias de férias. I hope that he considers (past subj.) the desires of the public and that he does not take (pres. subj.) so many vacation days.

Acceptability Task

The acceptability task consisted of 10 sentences which learners were asked to rate as

either correct or incorrect. Results for this task are shown in Figure 6:

Figure 6: Accuracy for acceptability task.

01020304050607080

High Beg Interm Low Adv

%

HLFL

As we can see in Figure 6, HL learners performed better in this task than NHL learners in every

level. We also see that NHL learners performed progressively better as the level of instruction

rose. This was not the case for HL learners: those enrolled in intermediate classes performed a

little better than those taking low advanced classes (a difference of roughly 2%). Note, too, that

in the low advanced level the difference between HL and NHL learners is also very small, at less

than 2%.

In this section, results were presented for the four tasks completed by participants in this

study. We saw that, overall, HL learners perceive and produce the subjunctive more accurately

than do NHL learners. The next section presents a discussion of these findings as they relate to

the questions posited in the beginning of this paper.

16

Discussion

The first goal of this investigation was to determine whether heritage language learners of

Portuguese perceived and produced subjunctive forms more accurately than foreign language

learners. From the results presented in the previous section, we can conclude that HL learners did

in fact perform better in the tasks at hand than did NHL learners. The difference between the two

groups did not always appear to be very significant, however. The largest differences between

the two groups of learners were attested in relation to the multiple choice task. For the other

tasks, the differences found were smaller. This finding is somewhat consistent with those

described in Fairclough (Spanish), although in that study larger differences were found between

accuracy of forms produced by HL and NHL learners. Lynch (“Similarities”) did not find a large

difference in the production of HL and NHL learners of Spanish, but the participants in his study

were lower-proficiency learners. In this study, the tendency is for the difference between HL and

NHL learners to become smaller as the level of proficiency rises, not the opposite.

The second question addressed in this paper regards length of instruction and whether

learners enrolled in more advanced levels perceive and produce the subjunctive more accurately

than those enrolled in less advanced classes. The answer to this question seems to be affirmative

for the NHL group in every task. Heritage language learners, on the other hand, do not

necessarily perform better with longer instruction. It is true that the scores for the HL learners in

the high beginner level were always lower than for HL learners in the higher levels. However,

HL learners enrolled in intermediate level responded more accurately (even if by small margins)

than HL learners in low advanced classes in the acceptability task and in the paragraph task. So,

length of instruction appears to correlate to performance only weakly when it comes to the HL

17

learners in this study. Furthermore, as noted above, the differences between the HL and the NHL

groups in most tasks become smaller as length of instruction increases.

The final goal of this study was to find out whether there is some correlation between

self-assessment and actual production of accurate forms of the subjunctive. Learners enrolled in

lower levels displayed less confidence in their abilities, which correlated positively with the

results of the tasks (learners in high beginner classes did not perform as well as those in

intermediate/low advanced levels). Even though, overall, both HL and NHL learners appeared to

be somewhat confident in their abilities regarding the subjunctive (as assessed by Statement (g)),

NHL learners in low advanced classes were even more confident than their HL counterparts and

than NHL learners in other levels: every NHL learner in low advanced classes agreed with

Statement (g). Although they did not display perfect accuracy in the tasks, the low advanced

NHL group did perform well and very close to their peers in the HL group. Therefore, there

appears to be some correlation between self-assessment and perception/production accuracy for

participants in this study.

Since HL learners perform more accurately than NHL learners already in lower levels,

differentiated instruction may be in the best interest of both groups of learners. With instruction

designed for heritage language learners, this group might be able to increase accuracy more

quickly than they normally do and might not display the slight decrease found between the

intermediate and the low advanced groups in some cases. With the current type of instruction,

which aims at foreign language learners (including the textbooks used in the classes surveyed),

heritage language learners may not have the opportunity to develop standard forms of the

subjunctive mood as rapidly or as accurately as they might if classes targeted their specific

needs. As Lynch points out, HL learners in traditional foreign language classes “spend extensive

18

amounts of time formally studying some basic elements of the language that they already control

and not giving adequate attention to other basic elements that they do not” (“Similarities” 274).

Moreover, explicit grammar instruction, which is typical (and perhaps needed) in NHL classes,

may not be so useful for heritage language learners: Schwarzer and Petrón report that the

heritage learners of Spanish in their study found the grammatical focus in their classes to be of

little use to them. On the other hand, Hislope calls for more explicit grammar instruction with a

focus-on-form activity for heritage speakers.

In terms of language teaching, there may be several ways to help learners develop

grammatical accuracy. Researchers such as Ellis and Fairclough (Spanish) have called for use of

awareness-raising activities. These activities may include task and test correction, where learners

have the opportunity to analyze their own language, individually or in groups (if carried out

collectively, the errors must remain anonymous, of course). Learners may also interview native

speakers, transcribe the interviews and identify/describe the cases when the subjunctive is used.

HL instruction would focus on the form that is appropriate in a given grammatical and

sociolinguistic context. This approach would be in line with what Parodi suggests for grammar

instruction for heritage learners. She argues that grammar should be presented as an analytical

subject, since the process of learning grammar is not the same for HL learners as it is for foreign

language learners: “While foreign language students employ grammar to have access to the

language, heritage speakers use the language to have access to the grammar” (211).

The goal of HL instruction is to teach standard dialects without losing sight of the

importance of learners’ home dialect. As highlighted by Fairclough (Spanish), Valdés (“The

teaching”), Lynch (“Similarities”), and others, HL teaching needs to build on the linguistic

knowledge and social experience that learners bring. These same authors also call attention to the

19

importance of socio-affective factors, arguing that negative attitudes toward the learners’ home

dialect can hinder the learning process. Therefore, in what regards heritage language teaching, it

is important to contrast and compare the home dialect and the standard dialect, in order to

facilitate the separation of the two. Fairclough (Spanish 136) also reminds us that teaching a

different dialect is a lengthy process that should be carried out over several semesters.

Concluding Remarks

This paper reports on investigation related to the written production of learners of

Portuguese in what regards the subjunctive vs. the indicative mood. Participants were divided

into heritage and foreign language learners, in order to assess whether the two groups exhibited

differences in grammatical accuracy. Results point to differences between the two groups of

learners regarding perception and production of the Portuguese subjunctive. These differences

decrease with length of instruction: the more advanced groups in this study displayed results that

were rather close in percentage terms. The smaller difference between the two groups of learners

may be interpreted in two different ways. On one hand, it may suggest that NHL learners “catch

up”, as it were, with HL learners over the course of instruction. On the other hand, the smaller

difference between the two groups in more advanced levels may suggest that HL learners do not

have the opportunity to access their knowledge of grammar, since they have the same kind of

instruction provided to NHL learners. Following authors such as Valdés (“Implications”), Parodi,

Lynch (“Similarities”), and many others, it has been suggested that instruction be differentiated

for HL and NHL learners. Since HL learners tend to possess some knowledge of the subjunctive

from very early on, differentiated instruction may help HL learners build on what they already

know. After all, “traditional” grammar lessons may be of little use to these learners, as argued by

20

Parodi (206) and Schwarzer and Petrón (574). Furthermore, in teaching heritage language

learners, it is imperative that instructors keep in mind the importance of socio-affective factors

and of the learners’ home dialect. Heritage learners of Portuguese should be exposed to different

registers of the language, which would allow for some experimentation with the language used in

various situations (Schwarzer and Petrón 577).

This study has contributed to the incipient field of Portuguese heritage language learning.

However, even though it included four different written tasks, it did not include any oral

production tasks. Future studies may compare the oral production of HL and NHL learners to

confirm (or not) the findings presented here. Furthermore, other studies may include a larger

percentage of NHL learners, since only less than 30% of the participants in this project belonged

to this group. Other linguistic features (not only grammatical, but also pragmatic, sociolinguistic,

discursive, etc.) should also be investigated in order to determine to what extent the production

of NHL and HL learners of Portuguese differs, and to what extent it is the same. With this type

of data, we may be able to design courses that are better suited for both groups of students.

21

Works Cited

Acevedo, Rebeca. “Navegando a través del Registro Formal.” Mi lengua: Spanish as a heritage language in the United States. Eds. A. Roca and M. C. Colombi. Washington, D.C.: Georgetown University Press, 2003. 257-268.

Ahern, Aoife and Manuel Leonetti. “The Spanish Subjunctive: Procedural Semantics and

Pragmatic Inference.” Current Trends in the Pragmatics of Spanish. Eds. R. Márquez-Reiter and M. E. Placencia. Amsterdam: John Benjamins, 2004. 35-56.

Bernal-Enríquez, Ysaura and Eduardo Hernández Chávez. “La Enseñanza del Español en Nuevo

México: ¿Revitalización o Erradicación de la Variedad Chicana?” Mi Lengua: Spanish as a Heritage Language in the United States. Eds. A. Roca and M. C. Colombi. Washington, D.C.: Georgetown University Press, 2003. 96-119.

Blake, Robert. “Mood Selection among Spanish-Speaking Children, Ages 4 to 12.” The

Bilingual Review/La Revista Bilíngüe 10 (1983): 21-32. Bull, William. The Visual Grammar of Spanish. Regents of the University of California and

Houghton Mifflin, 1972. Ellis, Rod. “Grammar Teaching – Practice or Consciousness Raising?” Methodology in

Language Teaching. Eds. J. Richards and W. Renandya. Cambridge: Cambridge University Press, 2002. 167-174.

Fairclough, Marta. “La Adquisición del Dialecto Estándar: Un Estudio Longitudinal del Discurso

Hipotético en el Español de Houston.” Contactos y Contextos Lingüísticos: El Español en los Estados Unidos y en Contacto con Otras Lenguas. Eds. L. Ortiz-López and M. Lacorte. Madrid: Iberoamericana, 2004. 161-172

Fairclough, Marta. Spanish and Heritage Language Education in the United States: Struggling

with Hypotheticals. Madrid: Iberoamericana, 2005. Ferreira, Fernanda. “That’s Not How My Grandmother Says It: Portuguese Heritage Learners in

Southeastern Massachusetts.” Hispania 88 (2005): 848-862. Ferreira, Fernanda. “Portuguese Heritage Language Learners: Proficiency Levels and

Sociolinguistic Profiles.” Portuguese Language Journal 2 Fall 2007. 15 November 2007. http://www.latam.ufl.edu/portugueselanguagejournal/index.html.

Gragera, Antonio. “The Role of Typological Markedness in the Acquisition of Spanish

Subjunctive and in Language Change.” Diss. University of Massachusetts Amherst, 2000.

22

Hislope, Kristi. “A Present Subjunctive Focus-on-Form Study of Heritage Speakers of Spanish.” Contactos y Contextos Lingüísticos: El Español en los Estados Unidos y en Contacto con Otras Lenguas. Eds. L. Ortiz-López and M. Lacorte. Madrid: Iberoamericana, 2004. 315-322.

Kondo-Brown, Kimi. “Differences in Language Skills: Heritage Language Learner Subgroups

and Foreign Language Learners.” The Modern Language Journal 89 (2005): 563-581. ______. Ed. Heritage Language Development: Focus on East Asian Immigrants.

Amsterdam/Philadelphia: John Benjamins, 2006. ______ and James D. Brown. Eds. Teaching Chinese, Japanese, and Korean Heritage Language

Students: Curriculum Needs, Materials, and Assessment. New York and London: Lawrence Erlbaum Associates, 2008.

Jouët-Pastré, Clémence. “Community-Based Learning: A Window into the Portuguese-Speaking

Communities of New England.” Hispania 88 (2005): 863-872. Lunn, Patricia. “Spanish Mood and the Prototype of Assertability.” Linguistics 27 (1989): 687-702.

Lynch, Andrew. “The Relationship between Second and Heritage Language Acquisition: Notes

on Research and Theory Building.” Heritage Language Journal 1.1 (2003). 10 January 2008. http://www.heritagelanguages.org/

Lynch, Andrew. “The Linguistic Similarities of Spanish Heritage and Second Language

Learners.” Foreign Language Annals 41 (2008): 252-281. Parodi, Claudia. “Stigmatized Spanish Inside the Classroom and Out.” Heritage Language

Education: A New Field Emerging. Eds. D. Brinton, O. Kagan, & S. Bauckus. London and NY: Routledge, 2007. 199-214.

Perini, Mário. Modern Portuguese: A Reference Grammar. New Haven: Yale University Press,

2001. Roberts, Ian and Mary Kato. Eds. Português Brasileiro: Uma Viagem Diacrônica. Campinas:

Editora da Unicamp, 1993. Santos, Denise and Gláucia V. Silva. “Authenticating Materials through Critical Thinking: The

Case of Teaching and Learning Suggestions in Portuguese.” Hispania 91 (2008): 110-123.

Schwarzer, David and Maria Petrón. “Heritage Language Instruction at the College Level:

Reality and Possibilities. Foreign Language Annals 38 (2005): 568-578. Silva-Corvalán, Carmen. “Current Issues in Studies of Language Contact.” Hispania 73 (1990):

162-176.

23

Stephens, Thomas. “Language Maintenance and Ethnic Survival: The Portuguese in New

Jersey.” Hispania 72 (1989): 716-720. Tse, Lucy. “Affecting Affect: The Impact of Ethnic Language Programs on Student Attitudes.”

Heritage Language Development. Eds. S. Krashen, L. Tse and J. McQuillan. Culver City, CA: Language Education Associates, 1998. 51-72

Valdés, Guadalupe. “Pedagogical Implications of Teaching Spanish to the Spanish-Speaking in

the United States.” Teaching Spanish to the Spanish Bilingual: Issues, Aims and Methods. Eds. G. Valdés, A. Lozano and R. García-Moya. New York: Teachers College, 1981. 3-20.

______“The Teaching of Spanish to Bilingual Spanish-Speaking Students: Outstanding Issues

and Unanswered Questions.” La Enseñanza del Español a Hispanohablantes: Praxis y Teoría. Eds. M.C. Colombi and F. Alarcón. Boston. Houghton Mifflin, 1997. 8-44.

______, Anthony Lozano and Rodolfo García-Moya. Eds. Teaching Spanish to the Spanish

Bilingual: Issues, Aims and Methods. New York: Teachers College, 1981. Wherritt, Irene. “A Sociolingüística da Oração Substantiva em Português.” Língua e Literatura 7

(1978): 85-109.

24

Appendix A

1. When you were growing up, what language(s) did you speak at home?

______________________________________________________________

2. When you were growing up, what language(s) did your relatives speak at home?

______________________________________________________________

3. Portuguese class(es) you are taking: _________________________________

4. Please rate the following statements regarding your perception of grammar in Portuguese.

Please be honest in your answers; they will NOT affect your class standing.

Agree Neither agree

nor disagree

Disagree

a. I know Portuguese grammar well.

b. I think “correct” grammar is not essential for people to

understand each other.

c. I have problems with different verb tenses in Portuguese.

d. The subjunctive is the most difficult verb form in

Portuguese.

e. I speak Portuguese better than I write it.

f. I write Portuguese better than I speak it.

g. I know when to use the subjunctive in written language.

h. I know when to use the subjunctive in spoken language.

Your Number

25

Appendix B Responde às questões abaixo em português. A- Escolhe a forma correcta do verbo: 1- Duvido que eles a) concordem com isso. b) concordam c) concordariam 2- Mesmo que ela a) venha nós vamos ao cinema. b) vem c) vier 3- Talvez a) vou visitar-te este fim-de-semana. b) vaia c) vá 4- Se a) ganhes a lotaria, compras uma casa. b) ganhares c) ganhasses 5- Era necessário que a) consigamos ultrapassar o problema. b) conseguíssemos c) conseguirmos 6- A não ser que a) encontro outro livro melhor, levo este. b) encontre c) encontrasse 7- Quando a) cheguei a casa, telefono-te. b) chegar c) chegue 8- Se a) quiseres, ia buscar-te. b) quisesses c) queres B. Completa com a forma correcta entre parêntesis: O Pedro é habitualmente uma pessoa calma. (1) (Gostar) _____________ do silêncio, de uma música suave e de ler... (2) (Ler) ____________ muito, em casa. Mas a irmã (3) (ser)__________ exactamente o contrário e por isso às vezes eles (4) (irritar-se)_________ um com o outro. Ela quer que ele (5) (sair) ______________, que se (6) (divertir) ________________, (7) (ir) ___________ ao cinema, (8) (ouvir) _______________ música rock, (9) (participar) _________________ em festas, sei lá... Mas ele não (10) (ser) __________ capaz e (11) (tentar) _____________ que ela (12) (compreender) __________________ que os

Your Number

26

gostos dele não (13) (ser) ______________ os dela. E (14) (querer) _________________ que ela (15) (saber) _________________ que ele (16) (ser) __________ mais feliz assim. Mas parece ser impossível que ela (17) (ser) ___________ tão feliz como ele... Ele não (18) (duvidar) _____________ que a felicidade (19) (estar) ___________ dentro das pessoas mas a irmã ainda não (20) (descobrir) ________________ e talvez não (21) (descobrir) ________________ nunca. Ela só (22) (saber) _______________ andar à velocidade dos outros... O Pedro tem agora cinco dias de férias e está a pensar ir ao norte do país se não (23) (chover) _____________ e não (24) (fazer) _____________ muito frio. É sempre imprevisível saber o tempo no Inverno. Por isso o Pedro (25) (decidir) _____________ telefonar para os Serviços de Meteorologia; se eles (26) (dizer) ___________ que (27) (fazer) ___________ sol ele (28) (ir) ___________ ao norte; se o tempo (29) (estar) ____________ mau (30) (ficar) ____________ em casa. C. Escreve um breve parágrafo sobre o seguinte tema: O que esperas do próximo Presidente. ______________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ D. Na tua opinião, as frases abaixo são correctas ou incorrectas? Marca com um C as que consideras correctas e com um I as que consideras incorrectas.

1. ______ Eu estudava mais se tinha tempo. 2. ______ Quando eu for a Portugal, vou ao Algarve. 3. ______ Se eu vá a Portugal, ficarei em Lisboa. 4. ______ Espero que a Maria está em Coimbra. 5. ______ Parece que a Maria está em Coimbra. 6. ______ Telefona-me quando chegarás. 7. ______ Sempre que puderes, vem visitar-me. 8. ______ Amanhã vamos à praia a menos que chova. 9. ______ Mesmo que faz sol, o Carlos vai ficar em casa. 10. ______ Se estivermos em casa, assistiremos a novela.

27

Notes

*I am greatly indebted to Sandra Sousa for her invaluable help with data collection and

tabulation. I would also like to thank the learners who participated in this study. Finally, thanks

are due to the anonymous reviewers who read a previous version of this paper. Needless to say,

all shortcomings are my own.

1 Note that the linguistic situation in Cape Verde is far from simple. In that archipelago,

Portuguese is the official language and is learned in schools and is heard on radio and television.

However, the language of oral communication in everyday life is Capeverdean Creole. The

linguistic issues in Cape Verde fall outside of the scope of this paper.

2 Heritage (and many native) speakers of European Portuguese in the American Northeast

use sinó for “snow” (Port neve), estoa for “store” (Port loja), televeijo for “television” (Port

televisão), and many other calques. Speakers of Brazilian Portuguese residing in the U.S. may

use bisado for “busy” (Port ocupado), parquear for “to park” (Port estacionar), printar for “to

print” (Port imprimir).

3 Note that (at least certain) dialects of Brazilian Portuguese display a tendency to

substitute the present subjunctive with the present indicative, as attested by Wherritt (1978) and

several studies in Roberts and Kato (1993). This tendency does not apply to the past and future

subjunctive. To the best of my knowledge, European Portuguese does not exhibit the same

tendency.

4 Note that the present indicative in the subordinate clauses in (2b) and (2c) would only

be possible if the verbs in the main clauses also appeared in the present indicative:

(i) Se ela vai à reunião, nós chegamos a um acordo.

28

If she goes to the meeting, we reach an agreement.

(ii) Quando ela vai à reunião, nós chegamos a um acordo.

When she goes to the meeting, we reach an agreement.

5An anonymous reviewer considers this a common mistake among NHL leaners.

According to the reviewer, the error may occur even before students learn the personal infinitive.

The reviewer considers that it may be related to not remembering the irregular past tense

conjugation of dizer.